ADCI (5-aminocarbonyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10- imine), a low-affinity uncompetitive NMDA antagonist, is a broad-spectrum anticonvulsant with a favorable side effect profile. However, the drug's clinical utility will be dependent upon its ability to maintain efficacy when it is administered to patients on a chronic basis. Therefore, we sought to determine if tolerance develops to the anticonvulsant activity of ADCI using the maximal electroshock (MES) test to assess seizure protection. Mice were treated with three daily injections of a 2 - ED50 dose for MES protection (18 mg/kg, i.p.) or vehicle for 7 or 14 days. On the day after the chronic treatment protocol, all animals received a challenge dose of ADCI (18 mg/kg) and 15 min later were evaluated in the MES test. In control animals, 83-94% of animals were protected and the ADCI plasma levels immediately after the MES test were 5.5-9.7 microg-ml. In treated animals, 29% and 0% of animals were protected at 7 and 14 days, respectively, and the ADCI plasma levels were 77% and 52% of control. [3H]Dizocilpine binding to brain NMDA receptors was unaltered by the chronic drug treatment. In subsequent experiments, we determined that 14-day chronically treated animals could be completely protected by increased doses of ADCI (ED50, 28.9 mg/kg). In both naive and chronically treated animals receiving a challenge dose of ADCI, ADCI plasma levels declined in two phases, the first with time constant of about 55 min and the second with a much slower rate. The estimated plasma concentrations of ADCI reflecting threshold (3-5 microg-ml) and 50% protection (5-7.5 microg/mg) were similar in naive and chronic animals. We conclude that tolerance to ADCI is due to pharmacokinetic factors (enhanced first-pass metabolism) and does not result from a reduction in anticonvulsant efficacy.